Neutron reflectivity surface layer thickness

The polymer concentration profile has been measured by small-angle neutron scattering from polymers adsorbed onto colloidal particles [70,71] or porous media [72] and from flat surfaces with neutron reflectivity [73] and optical reflectometry [74]. The fraction of segments bound to the solid surface is nicely revealed in NMR studies [75], infrared spectroscopy [76], and electron spin resonance [77]. An example of the concentration profile obtained by inverting neutron scattering measurements appears in Fig. XI-7, showing a typical surface volume fraction of 0.25 and layer thickness of 10-15 nm. The profile decays rapidly and monotonically but does not exhibit power-law scaling [70]. 

The use of neutron reflectivity at liquid interfaces, which is a method sensitive to both surface roughness and surfactant layer thickness, was reviewed with the examples of polydimethylsiloxane-surfactant layers.633 Sum-frequency generation (SFG) vibrational spectroscopy was applied to study surface restructuring behavior of PDMS in water in an attempt to understand antifouling properties of silicones.6 ... 

The effect of constraints introduced by confining diblock copolymers between two solid surfaces was examined by Lambooy et al. (1994) and Russell et al. (1995). They studied a symmetric PS-PMMA diblock sandwiched between a silicon substrate, and silicon oxide evaporated onto the top (homopolymer PMMA) surface. Neutron reflectivity showed that lamellae formed parallel to the solid interfaces with PMMA at both surfaces. The period of the confined multilayers deviated from the bulk period in a cyclic manner as a function of the confined film thickness, as illustrated in Fig. 2.60. First-order transitions were observed at t d0 = (n + j)d0, where t is the film thickness and d0 is the bulk lamellar period, between expanded states with n layers and states with (n + 1) layers where d was contracted. Finally, the deviation from the bulk lamellar spacing was found to decrease with increasing film thickness (Lambooy et al. 1994 Russell et al. 1995). These experimental results are complemented by the phenomenologi-... 

The more recent neutron reflectivity studies have established that flattened surface micelle or fragmented bilayer structure in more detail and with more certainty, using contrast variation in the surfactant and the solvent [24, 31]. However, the extent of the lateral dimension (in the plane of the surface) and the detailed structure in that direction is less certain. From those neutron reflectivity measurements [24, 31] and related SANS data on the adsorption of surfactants onto colloidal particles [5], it is known that the lateral dimension is small compared with the neutron coherence length, such that averaging in the plane is adequate to describe the data. The advent of the AFM technique and its application to surfactant adsorption [15] has provided data that suggest that there is more structure and ordering in the lateral direction than implied from other measurements. This will be discussed in more detail in a later section of the chapter. At the hydrophobic interface, although the thickness of the adsorbed layer is now consistent with a monolayer, the same uncertainties about lateral structure exist. 

An alternative approach to producing a differently functional surface is to use spin coating techniques. This was done by Turner et al. [68], who spun cast layers of polystyrene onto a silica surface. They investigated the nature of the surface and of SDS adsorbed to that surface by neutron reflectivity and IR-ATR. A thin layer of polystyrene, 275 A, was established. The subsequent SDS adsorption was consistent with a monolayer 15 A thick and an adsorbed amount similar to that observed at the air-solution interface. Measurements above the cmc of SDS showed clearly the effects on the adsorption pattern of dodecanol impurities in the SDS. 

A feature of the neutron reflectivity study on polyDMDAAC and surfactant adsorption by Penfold et al. [74] was that the adsorbed layer of polyDMDAAC was remarkably robust and unaffected by the subsequent surfactant adsorption. This is not always the case, and Fielden et al. [76] reported a large increase in the thickness of the surface layer of AM-MAPTC on mica due to complex formation with SDS. Thickness increases with electrolyte and pH were reported for high molecular weight polyacrylamide adsorbed onto silica, measured by null ellip-sometry by Samoshina et al. [82] in the absence of surfactant. Complex formation at the interface, resulting in layer thickening, was also reported by Dedinaite et al. [83] for PCMA/SDS mixtures on mica from AFM measurements. 

Extensive neutron reflectivity studies on surfactant adsorption at the air-water interface show that a surfactant monolayer is formed at the interface. Even for concentration cmc, where complex sub-surface ordering of micelles may exist,the interfacial layer remains a monolayer. This is in marked contrast to the situation for amphiphilic block copolymers, where recent measurements by Richards et al. on polystyrene polyethylene oxide block copolymers (PS-b-PEO) and by Thomas et al. on poly(2-(dimethyl-amino)ethylmethacrylamide-b-methyl methacrylate) (DMAEMA-b-MMA) show the formation of surface micelles at a concentration block copolymer, where an abrupt change in thickness is observed at a finite concentration, and signals the onset of surface micellisation. 

Fig. 12. a Neutron reflectivity data for a 860 A-thick PMDA-3F/PPO triblock film which has been foamed for 4 h. The line through the data is the fit using the scattering density profile shown in b. b Scattering length density profiles for a unfoamed and foamed PMDA-3F/PPO triblock. The foaming results in a decrease in the overall film thickness and an increase in the density of the skin layer, particularly at the air interface. The density in the center of the film corresponds to about 15% voids, density skin at the top and bottom surfaces of the film... 

The significant increase in co, with increasing m is caused by the localisation of the oxyethylene chain in the surface layer. This result which is implied by the reorientation model also agrees with the neutron reflection data. It was shown in [14] that, with the increase of the area per C EO , molecule in the adsoiption layer, i.e., with surface pressure deerease, the thickness of the layer occupied by the oxyethylene groups of C EO, becomes lower. At the same time, a decrease in the tilt angle of the oxyethylene groups to the interface is observed. These results were discussed in [14] in the context of the adsorption of oxyethylene groups in the non-saturated adsorption layer of oxyethylated alcohols. The dependence of molar areas in the two states on n is shown in Fig. 3.32. 

Studies have been done to investigate the orientation and adsorption of surface active dyes at the oil/water interface using fluorescence [74, 75], resonance Raman scattering [76, 77]. Neutron reflectivity has been used [78] to determine the thickness of a surfactant (monodecyl tetraglycol ether) layer at the octane/water interface and the interfacial roughness, which was large (m. 90 A) due to the very low interfacial tension (7 = 0.08 mN rn ). 

Three direct methods can be applied for determination of adsorbed layer thickness ellipsometry, attenuated total reflection (ATR) and neutron scattering. The first two [38] depend on the difference between refractive indices between the substrate, the adsorbed layer and bulk solution and require a flat reflecting surface. Ellipsometry [38] is based on the principle that light undergoes a change in polarizability when it is reflected at a flat surface (whether covered or uncovered with a polymer layer). 

Methods based on the neutron scattering give the concentration profile and the second moment of the profile, i.e., the mean square thickness of the layer. A neutron reflection is also very useful for these studies. A comprehensive and coherent series of neutron reflectivity data of pol3rmers adsorbed at the surface has been presented. The data allow one to obtain the distribution function, 0(z), using some realistic hypothesis, including independence of 0(z)... 

---